Optimised State of Charge Estimation for Battery Using ML Technique

Correct estimation of State of Charge in electric vehicles, i.e., critical step for better performance, longer lifespan, and safety in batteries. Traditional methods of SOC estimation using utilization of Coulomb counting and voltage-based methods are not accurate keeping in view environmental conditions as well as aging of the batteries. Efficient SOC estimation using machine learning techniques along with BMS over ESP8266 microcontroller is proposed in this work. It relies on ML mechanisms for real-time accurate SOC estimation based on cumulative battery data as well as providing efficient measurement of data and wireless communication through ESP8266 for remote monitoring as well as supervision. The proposed system is aimed towards increased reliability as well as efficiency of SOC estimation for better performance of batteries, safe operations, and higher mileage of vehicle. The main advantage of this technique is the ability to overcome traditional methods' limitation, hence providing better precision in real-time SOC.